Association of MRI Measures With Disease Severity and Progression in Progressive Supranuclear Palsy

Abstract

Objective: To verify the association of midbrain-based MRI measures as well as cortical volumes with disease core features and progression in patients with Progressive Supranuclear Palsy (PSP). Methods: Sixty-seven patients (52.2% with Richardson's syndrome) were included in the present analysis. Available midbrain-based MRI morphometric assessments as well as cortical lobar volumes were computed. Ocular, gait and postural involvement at the time of MRI was evaluated with the PSP rating scale. Specific milestones or death were used to estimate disease progression up to 72 months follow up. Hierarchical regression models and survival analysis were used for analyzing cross-sectional and longitudinal data, respectively. Results: Multivariate models showed vertical supranuclear gaze palsy was associated with smaller midbrain area (OR: 0.02, 95% CI 0.00-0.175, p = 0.006). Cox regression adjusted for age, disease duration, and phenotype demonstrated that lower midbrain area (HR: 0.122, 95% CI 0.030-0.493, p = 0.003) and diameter (HR: 0.313, 95% CI 0.112-0.878, p = 0.027), higher MR Parkinsonism Index (HR: 6.162, 95% CI 1.790-21.209, p = 0.004) and larger third ventricle width (HR: 2.755, 95% CI 1.068-7.108, p = 0.036) were associated with higher risk of dependency on wheelchair. Conclusions: Irrespective of disease features and other MRI parameters, reduced midbrain size is significantly associated with greater ocular motor dysfunction at the time of MRI and more rapid disease progression over follow up. This is the first comprehensive study to systematically assess the association of available midbrain-based MRI measures and cortical volumes with disease severity and progression in a large cohort of patients with PSP in a real-world setting.

Keywords: disease progression; disease severity; imaging; milestones; progressive supranuclear palsy.

Copyright © 2020 Picillo, Abate, Ponticorvo, Tepedino, Erro, Frosini, Del Prete, Cecchi, Cosottini, Ceravolo, Salle, Salle, Esposito, Pellecchia, Manara and Barone.

Figures

Figure 1

Overview of the midbrain-based measures computed. Sagittal (A,B,F), coronal (C) and axial (D,E,G,H) T1-weighted volumetric MR images. Midsagittal midbrain area (1) and midsagittal pons area (2) used for calculating the pons-to-midbrain area ratio (P/M) are shown in (A). Middle and superior cerebellar peduncles used to compute middle cerebellar peduncles to superior cerebellar peduncles ratio (MCP/SCP) are shown in (B) and (C), respectively. Third ventricle measurements at the level of anterior and posterior commissures [third ventricle width derives from the mean of the anterior (1), medium (2), and posterior lines (3)] and frontal horns width are shown in (D) and (E), respectively. MRPI is calculated with the following formula: (P/M) × (MCP/SCP); MRPI 2.0 is calculated with the following formula: MRPI × (third ventricle width/frontal horns width); P/M 2.0 is calculated with the following formula: (P/M) × (third ventricle width/frontal horns width). Midbrain and pons diameters obtained from midsagittal elliptical regions of interests are shown in (F). Minor axes were used to calculate the pons-to-midbrain diameter ratio. The interpeduncular angle, calculated on a plane parallel to anterior commissure-posterior commissure line and right below the mammillary bodies, is shown in (G). The length of midbrain tegmentum, measured as the distance between the interpeduncular fossa and the center of the aqueduct at the level of mammillary bodies, on a plane parallel to the anterior commissure-posterior commissure line is shown in (H). MCP/SCP, middle cerebellar peduncles to superior cerebellar peduncles ratio; MR, magnetic resonance; MRPI, MR Parkinsonism index; MRPI 2.0, MR Parkinsonism index 2.0 version; P/M, pons-to-midbrain area ratio; P/M 2.0, pons-to-midbrain area ratio 2.0 version.

Figure 2

Cumulative risk of dependency on wheelchair by each MRI parameter (less atrophy vs. more atrophy). Only curves with significant log-rank test at 0.05 level are reported. Dotted line: values above median; Continuous line: values below median.

Figure 3

Cumulative risk of death by each MRI parameter (less atrophy vs. more atrophy). Only curves with significant log-rank test at 0.05 level are reported. Dotted line: values above median; Continous line: values below median.